Underreporting and missed opportunities for uptake of intermittent preventative treatment of malaria in pregnancy (IPTp) in Mali

Objectives: To identify factors contributing to low uptake of intermittent preventive treatment of malaria in pregnancy with sulfadoxine-pyrimethamine (IPTp-SP) in rural Mali. Methods: We conducted secondary data analysis on Mali’s 2012-2013 Demographic and Health Survey (DHS) to determine the proportion of women who failed to take IPTp-SP due to ineligibility or non-attendance at antenatal care (ANC). We also identified the proportion who reported taking other or unknown medications to prevent malaria in pregnancy and those who did not know if they took any medication to prevent malaria in pregnancy. We conducted qualitative interviews, focus groups and ANC observations in six rural sites in Mali’s Sikasso and Koulikoro regions to identify reasons for missed opportunities. Results: Our secondary data analysis found that reported IPTp-SP coverage estimates are misleading due to their dependence on a variable (“source of IPTp”) that is missing 62% of its data points. Among all women who gave birth in the two years prior to the survey, 56.2% reported taking at least one dose of IPTp-SP. Another 5.2% reported taking chloroquine, 1.9% taking another drug to prevent malaria in pregnancy, 4.4% not knowing what drug they took to prevent malaria, and 1.1% not knowing if they took any drug to prevent malaria. The majority of women who did not receive IPTp-SP were women who also did not attend ANC. Our qualitative data revealed that many health centers neither administer IPTp-SP by directly observed therapy, nor give IPTp-SP at one month intervals through the second and third trimesters, nor provide IPTp-SP free of charge. Women generally reported IPTp-SP as available and tolerable, but frequently could not identify its name or purpose, potentially affecting accuracy of responses in household surveys. Conclusion: We estimate IPTp-SP uptake to be significantly higher than stated in Mali’s 2012-13 DHS report. Increasing ANC attendance should be the first priority for increasing IPTp-SP coverage. Reducing cost and access barriers, ensuring that providers follow up-to-date guidelines, and improving patient counseling on IPTp-SP would also facilitate optimal uptake. Using the 2012–2013 Mali DHS maternal health dataset in Stata version 12 [23], we ran cross tabulations at each point where a woman could miss an opportunity to take or report having taken IPTp. First, we filtered out women who never attended ANC and those who attended only during their first trimester (prior to IPTp eligibility). Next, we examined the proportion of women who reported that they (a) took at least one dose of IPTp-SP, (b) took chloroquine as IPTp, (c) took an unknown drug to prevent malaria in pregnancy, (d) took another drug for to prevent malaria in pregnancy, (e) did not know if they took any drug to prevent malaria in pregnancy, and (f) did not take any drug to prevent malaria in pregnancy. Among women who reported having taken IPTp-SP1 and were eligible for a IPTp-SP2 (those who had two or more ANC visits after their first trimester), we identified the proportion who reported that they (a) took the second dose, (b) did not take the second dose, and (c) did not know if they took the second dose. To be consistent with the DHS report, we restricted analysis to women who gave birth in the two years prior to the survey. For each analysis, we examined the national sample, and the subset of women from rural areas. All statistics were weighted using DHS methodology [24]. We conducted 28 in-depth interviews (IDIs), 26 focus-group discussions (FGDs) and 29 ANC observations across six rural health zones across Sikasso and Koulikoro. These two regions offer advantages in the study design as they are geographically different from where similar studies have been conducted before (Segou) [11,12], yet have similar IPTp uptake estimates to the national averages [8]. We began by purposively selecting two rural health zones within each region, one served by a larger health center accessible by paved road in a central town or population center, and another served by a smaller health center deep in a rural village and far from a population center or paved road. Within these sites, we purposively sampled community members who could offer a range of perspectives on IPTp, including pregnant women, family members, health workers and community leaders (Table 2). * Numbers presented refer to number of focus groups, not number of participants. We determined we had reached data saturation in IDIs and FGDs with community members in the original four sites, as iterative analysis revealed recurring, developed themes both within and between sites [25,26]. However, because procedures for administering IPTp-SP differed significantly in each of the first four health facilities we observed, we decided to observe additional sites to better characterize that variability. Thus, we added one additional observation site per region. We also determined that themes regarding health system factors warranted triangulation from higher-level health officials. Thus, we added six key informant interviews with leaders at district- and regional-level reference hospitals toward the end of the study. The data collection team included four Malian women trained in qualitative research methods (one sociologist and three physicians) and the first author, an American doctoral student fluent in Bambara. IDIs and FGDs with community members were semi-structured and included topics on a) health, illness prevention and use of medications during pregnancy, b) experiences with ANC and IPTp-SP, and c) perceived barriers to IPTp-SP uptake. The original English versions of the IDI and FGD guides are included as an appendix (S1 Appendix). During key informant interviews, district and regional level health officials were asked about content and implementation of IPTp-SP policy and their views on factors contributing to uptake. The original English versions of the IDI and FGD guides are included as an appendix. IDIs and FGDs were conducted in Bambara, the local language, with the exception of some key informants and clinicians who preferred French. Participants were interviewed in private locations within the community (typically, the participant’s home or place of work). All IDIs and FGDs were audio-recorded, then translated and transcribed into French. The ANC observations were conducted in order to characterize the context and processes of IPTp-SP administration at each health center. Based on pre-testing, we determined that with two data collectors and a purposive sample five patients per center, we could produce a thorough and internally consistent description of that health center’s processes. The purposive sampling was intended to capture women at different points in their pregnancy. Data collectors observed each participant for the entirety of her ANC visit and completed a structured observation form covering the patient-provider interaction, clinical exam, content of patient counseling, and acquisition/administration of pharmaceuticals. Researchers also took unstructured notes detailing the events and context of the visit. Additionally, we reviewed maternity and pharmacy record books at each health center for SP stock and distribution information. Analysis of textual and observational data was an iterative, collaborative process. Interviewers wrote reflective memos after each data collection session and discussed developing themes at regular debriefing meetings [27]. Throughout data collection, we added and adjusted questions on IDI and FGD guides based on previous IDIs, FGDs, and debriefings. Transcripts and observation notes were uploaded into ATLAS.ti version 7 for descriptive coding [28]. The research team began coding inductively and periodically revised the codebook as new themes emerged. A sample of transcripts were double-coded to ensure consistency between researchers. Multiple members of the research team reviewed and discussed data outputs of each code to summarize findings. The Institutional Review Boards of the Johns Hopkins Bloomberg School of Public Health and the Faculty of Medicine, Pharmacy, and Odontostomatology at the University of Sciences, Techniques, and Technologies of Bamako approved the study. Informed consent was obtained from all participants of IDIs, FGDs and observations. Interviewers signed consent forms to document oral consent, which was collected in order to overcome literacy limitations and ensure participant anonymity.